Our laboratory has a longstanding interest in alterations in oncogenes, tumor suppressor genes, and DNA repair and cell cycle related genes that impact therapeutic efficacy and patient survival in patients with gastrointestinal malignancies. In this application, we focus on studies directed at understanding molecular alterations present in gastric cancer (GC). GCs cause 3.9 deaths per 100,000 persons each year in the U.S. and the prognosis has not changed in decades. Furthermore, proximal GCs are one of the most rapidly increasing cancers in the U.S. We postulate that the poor clinical outcome results from the presence of factors that promote disease progression by creating widespread genetic instability as well as the loss of proteins required to mediate programmed cell death (PCD). Further, we postulate that GC will exhibit microsatellite instability (MSI) with secondary mutations in the pro-apoptotic BAX gene. We also postulate that some GCs will show loss of heterozygosity (LOH) at the p53 and DCC loci negatively impacting patient outcome. "Molecular staging" could enhance TNM staging which is based solely on standard pathological variables. Questions to be answered include 1) Does an individual alteration have an impact on outcome, independent of TNM stage? 2) Does MSI or TGFBRII or BAX gene alteration induce therapeutic resistance? and 3) Is there a reliable measurement methodology for each marker? To examine these issues we propose the following specific aims: Specific Aim 1: To test the hypothesis that MSI is present in a significant number of GCs and when present, serves as an independent marker of patient prognosis and/or serves as a predictor of therapeutic resistance. Specific Aim 2: To test the hypothesis that mutations in the BAX gene occur in patients with MSI+ gastric cancers and that these BAX mutations negatively impact patient response to therapy. Specific Aim 3: To test the hypothesis that LOH occurs frequently in GCs and that loss of 17p (the p53 locus) and 18q (the DCC locus), independently predicts patient prognosis and possibly therapeutic response. The patients derive from a Southwest Oncology Group clinical protocol.